detailed look on lilypond output

[Michael Kiermaier]

Hello!

It is my strong believe that a perfect music typesetter only needs the pure 
musical information of a music file, without redundancy. For this reason I 
never was happy with Finale & Co., and for the very same reason lilypond 
looks tremendously appealing to me.

I finished my first lilypond a while ago and the result was overwhelming.
But still I see some points where I think that lilypond does not totally the 
right thing. Some of them went away when I updated from 2.0 to 2.2, but the 
bigger part remained.

I want to report and discuss them here.

Please download my file at
http://home.in.tum.de/~kiermaim/misc/Rag.ly
and compile it with lilypond-2.2.0

A)
bar 1: There is a line titled %HACK in the source file which is necessary to 
get the bar printed correctly. This is a bug documented in the manual. I was 
a bit surprised that 2.2 does not fix it.

B)
bar 1, left hand: The bass clef should be placed below the notes of the right 
hand. For example, in bar 17 lilypond sets the treble clef correctly. I am 
not sure why this does not happen in bar 1. 

C)
I feel that the accidentals of grace note are too much apart from the main 
note. This can be observed in many locations, the first one is the f# in the 
left hand of bar 6.

D)
bar 7, left hand: The bass clef is set too close to the preceeding bb-note.

E)
bar 9, left hand: There is a collision between the slur and the 3 of the 
triplet.

F)
bar 31, right hand: There is a collision of the grace note d and the ledger 
line of the succeeding chord.

G)
Many ties look too steep. For example in bar 32, right hand; or in bar 41, 
right hand. I am not sure if the horizontal alignment of the ties is correct. 
I looked at scores of Edition Peters, and I have the impression that ties get 
hooked horizontally to the center of the note heads, if this is possible 
without collision, otherwise they are hooked to the inner border (I hope this 
is understandable). Lilypond seems to hook always to the inner border.

H)
bar 37, right hand. There is a collision of two ties. In version 2.0 the tie 
of the note e went down, so there was no collision of the ties (but a 
collision with the ledger line). I guess the correct thing would be the 
behaviour of 2.0, but making the tie a bit flatter (or steeper?) to avoid the 
collision with the ledger line.

J)
bar 37, left hand. The upper slur collides with the flag. This clearly does 
not look good, but I am not sure what would be the correct way resolve it.

K)
bar 42, right hand. The ending of the 8va-bracket doesn't look good. It should 
end either clearly before or clearly behind the dot, in my opinion. In 2.0 it 
ended (too much) before the dot, but it did not look good, either.


Please note that I am not looking for manual fixes for these observations.
A, B, D, E, F, H, K look like bugs to me. Maybe I discovered some unknown 
ones. I hope that my post helps to fix them quickly.
Also I hope to have a discussion about the observations, especially the more 
arguable ones.


Many Thanks for this great program and keep up the good work,

~michael

-- 
"Mailand oder Madrid - Hauptsache Italien!"
-Andreas Möller